Darokrithia
commented
6 years ago I changed chrome_size in xdegnome to 15 and multiplied the initial values by i (so each value in data_array was i2 or i1 instead of 2 or 1). It seems like all the DNA comes from the dad, and there are two negative mutations. One of the second gene, and one on the fourth. This seems to suggest to me that the issue may be an RNG one. This also would fit with the earlier issue of the same thing happening but instead of being from the dad, all the DNA was based at zero. This is consistent as at that point I hadn't made the last piece of DNA get crossed over (e.g. all DNA before a crossover would be placed, and none after the final crossover). As the dad is the ending tip (from 0 to end), before fixing it would have given all zeros. Still I am not 100% trusting
Although everything runs fine (no segfaults), the majority of values generated with one generation runs seems plausible, and I have pretty diligently debugged everything, I don't trust the values. ./xdegnome seems to always give the same values, and they don't seem extremely believable. As such I think there is a clear issue either in rng implementation or more likely the Degnome_mate() function. Either way, until I am confident that the bugs are squashed, this issue will remain open.